fix some warnings from gcc and clang-tidy (#3038)

Co-authored-by: xaedes <xaedes@gmail.com>

12 files changed, 39 insertions, 65 deletions

diff --git a/examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp b/examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp
index 9e856c21..293b455d 100644
--- a/examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp
+++ b/examples/convert-llama2c-to-ggml/convert-llama2c-to-ggml.cpp
@@ -1,5 +1,6 @@
 #include "ggml.h"
 #include "llama.h"
+#include "common.h"
 
 #include <unordered_map>
 #include <vector>
@@ -499,10 +500,10 @@ struct llama_file {
         errno = 0;
         std::size_t ret = std::fread(ptr, size, 1, fp);
         if (ferror(fp)) {
-            throw std::runtime_error(format("read error: %s", strerror(errno)));
+            die_fmt("fread failed: %s", strerror(errno));
         }
         if (ret != 1) {
-            throw std::runtime_error(std::string("unexpectedly reached end of file"));
+            die("unexpectedly reached end of file");
         }
     }
 
@@ -597,8 +598,7 @@ void load_vocab(const char *filename, Config *config, struct llama_vocab *vocab)
         printf("Assuming llama2.c vocabulary since %s is not a gguf file\n", filename);
         llama_file file(filename, "rb");
         if (!file.fp) {
-            fprintf(stderr, "error: %s: %s\n", strerror(errno), filename);
-            exit(1);
+            die_fmt("%s: %s", strerror(errno), filename);
         }
         const int  n_vocab = config->vocab_size;
         /* uint32_t max_token_length =  */ file.read_u32(); // unused
diff --git a/examples/embd-input/embd-input-lib.cpp b/examples/embd-input/embd-input-lib.cpp
index 036bdb39..87aac347 100644
--- a/examples/embd-input/embd-input-lib.cpp
+++ b/examples/embd-input/embd-input-lib.cpp
@@ -23,7 +23,7 @@ extern "C" {
 struct MyModel* create_mymodel(int argc, char ** argv) {
     gpt_params params;
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return nullptr;
     }
 
diff --git a/examples/embedding/embedding.cpp b/examples/embedding/embedding.cpp
index 93d583b5..49ab3e06 100644
--- a/examples/embedding/embedding.cpp
+++ b/examples/embedding/embedding.cpp
@@ -11,7 +11,7 @@
 int main(int argc, char ** argv) {
     gpt_params params;
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
diff --git a/examples/gptneox-wip/falcon-main.cpp b/examples/gptneox-wip/falcon-main.cpp
index d4b130b2..7f9a1620 100644
--- a/examples/gptneox-wip/falcon-main.cpp
+++ b/examples/gptneox-wip/falcon-main.cpp
@@ -953,7 +953,7 @@ int main(int argc, char ** argv) {
 
     gpt_params params;
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
diff --git a/examples/gptneox-wip/gptneox-main.cpp b/examples/gptneox-wip/gptneox-main.cpp
index b6cc46c5..55eba0cd 100644
--- a/examples/gptneox-wip/gptneox-main.cpp
+++ b/examples/gptneox-wip/gptneox-main.cpp
@@ -925,7 +925,7 @@ int main(int argc, char ** argv) {
 
     gpt_params params;
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
diff --git a/examples/main/main.cpp b/examples/main/main.cpp
index 9201b53b..c9ca7719 100644
--- a/examples/main/main.cpp
+++ b/examples/main/main.cpp
@@ -48,8 +48,9 @@ static bool is_interacting = false;
 
 void write_logfile(
     const llama_context * ctx, const gpt_params & params, const llama_model * model,
-    const std::vector<llama_token> input_tokens, const std::string output, const std::vector<llama_token> output_tokens) {
-
+    const std::vector<llama_token> & input_tokens, const std::string & output,
+    const std::vector<llama_token> & output_tokens
+) {
     if (params.logdir.empty()) {
         return;
     }
@@ -109,7 +110,7 @@ int main(int argc, char ** argv) {
     gpt_params params;
     g_params = &params;
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
@@ -303,7 +304,7 @@ int main(int argc, char ** argv) {
 
     // debug message about similarity of saved session, if applicable
     size_t n_matching_session_tokens = 0;
-    if (session_tokens.size() > 0) {
+    if (!session_tokens.empty()) {
         for (llama_token id : session_tokens) {
             if (n_matching_session_tokens >= embd_inp.size() || id != embd_inp[n_matching_session_tokens]) {
                 break;
@@ -401,7 +402,7 @@ int main(int argc, char ** argv) {
 
         LOG_TEE("%s: interactive mode on.\n", __func__);
 
-        if (params.antiprompt.size()) {
+        if (!params.antiprompt.empty()) {
             for (const auto & antiprompt : params.antiprompt) {
                 LOG_TEE("Reverse prompt: '%s'\n", antiprompt.c_str());
             }
@@ -499,7 +500,7 @@ int main(int argc, char ** argv) {
 
     while ((n_remain != 0 && !is_antiprompt) || params.interactive) {
         // predict
-        if (embd.size() > 0) {
+        if (!embd.empty()) {
             // Note: n_ctx - 4 here is to match the logic for commandline prompt handling via
             // --prompt or --file which uses the same value.
             int max_embd_size = n_ctx - 4;
@@ -624,7 +625,7 @@ int main(int argc, char ** argv) {
                 LOG("n_past = %d\n", n_past);
             }
 
-            if (embd.size() > 0 && !path_session.empty()) {
+            if (!embd.empty() && !path_session.empty()) {
                 session_tokens.insert(session_tokens.end(), embd.begin(), embd.end());
                 n_session_consumed = session_tokens.size();
             }
@@ -695,7 +696,7 @@ int main(int argc, char ** argv) {
         // if not currently processing queued inputs;
         if ((int) embd_inp.size() <= n_consumed) {
             // check for reverse prompt
-            if (params.antiprompt.size()) {
+            if (!params.antiprompt.empty()) {
                 std::string last_output;
                 for (auto id : last_tokens) {
                     last_output += llama_token_to_piece(ctx, id);
@@ -732,7 +733,7 @@ int main(int argc, char ** argv) {
                 LOG("found EOS token\n");
 
                 if (params.interactive) {
-                    if (params.antiprompt.size() != 0) {
+                    if (!params.antiprompt.empty()) {
                         // tokenize and inject first reverse prompt
                         const auto first_antiprompt = ::llama_tokenize(ctx, params.antiprompt.front(), false);
                         embd_inp.insert(embd_inp.end(), first_antiprompt.begin(), first_antiprompt.end());
diff --git a/examples/perplexity/perplexity.cpp b/examples/perplexity/perplexity.cpp
index 843b2ae3..1b760683 100644
--- a/examples/perplexity/perplexity.cpp
+++ b/examples/perplexity/perplexity.cpp
@@ -655,7 +655,7 @@ int main(int argc, char ** argv) {
     gpt_params params;
 
     params.n_batch = 512;
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
diff --git a/examples/quantize-stats/quantize-stats.cpp b/examples/quantize-stats/quantize-stats.cpp
index 06ce18f0..6ce03ba7 100644
--- a/examples/quantize-stats/quantize-stats.cpp
+++ b/examples/quantize-stats/quantize-stats.cpp
@@ -71,7 +71,7 @@ void quantize_stats_print_usage(int /*argc*/, char ** argv) {
 }
 
 // Check if a layer is included/excluded by command line
-bool layer_included(const quantize_stats_params params, const std::string & layer) {
+bool layer_included(const quantize_stats_params & params, const std::string & layer) {
     for (const auto& excluded : params.exclude_layers) {
         if (std::regex_search(layer, std::regex(excluded))) {
             return false;
diff --git a/examples/quantize/quantize.cpp b/examples/quantize/quantize.cpp
index c174be06..1bf18248 100644
--- a/examples/quantize/quantize.cpp
+++ b/examples/quantize/quantize.cpp
@@ -143,10 +143,9 @@ int main(int argc, char ** argv) {
         if (!try_parse_ftype(argv[arg_idx], params.ftype, ftype_str)) {
             fprintf(stderr, "%s: invalid ftype '%s'\n", __func__, argv[3]);
             return 1;
-        } else {
-            if (ftype_str == "COPY") {
-               params.only_copy = true;
-            }
+        }
+        if (ftype_str == "COPY") {
+           params.only_copy = true;
         }
         arg_idx++;
     }
diff --git a/examples/save-load-state/save-load-state.cpp b/examples/save-load-state/save-load-state.cpp
index 573bc4ef..14e9501c 100644
--- a/examples/save-load-state/save-load-state.cpp
+++ b/examples/save-load-state/save-load-state.cpp
@@ -13,7 +13,7 @@ int main(int argc, char ** argv) {
     params.repeat_last_n = 64;
     params.prompt = "The quick brown fox";
 
-    if (gpt_params_parse(argc, argv, params) == false) {
+    if (!gpt_params_parse(argc, argv, params)) {
         return 1;
     }
 
@@ -44,7 +44,7 @@ int main(int argc, char ** argv) {
         llama_free_model(model);
         return 1;
     }
-    auto tokens = llama_tokenize(ctx, params.prompt.c_str(), true);
+    auto tokens = llama_tokenize(ctx, params.prompt, true);
     auto n_prompt_tokens = tokens.size();
     if (n_prompt_tokens < 1) {
         fprintf(stderr, "%s : failed to tokenize prompt\n", __func__);
diff --git a/examples/server/server.cpp b/examples/server/server.cpp
index 6b606447..3f3c6465 100644
--- a/examples/server/server.cpp
+++ b/examples/server/server.cpp
@@ -139,7 +139,7 @@ static std::string tokens_to_output_formatted_string(const llama_context *ctx, c
 }
 
 // convert a vector of completion_token_output to json
-static json probs_vector_to_json(const llama_context *ctx, const std::vector<completion_token_output> probs)
+static json probs_vector_to_json(const llama_context *ctx, const std::vector<completion_token_output> & probs)
 {
     json out = json::array();
     for (const auto &prob : probs)
@@ -271,7 +271,7 @@ struct llama_server_context
         return true;
     }
 
-    std::vector<llama_token> tokenize(json json_prompt, bool add_bos)
+    std::vector<llama_token> tokenize(const json & json_prompt, bool add_bos) const
     {
         // If `add_bos` is true, we only add BOS, when json_prompt is a string,
         // or the first element of the json_prompt array is a string.
@@ -611,7 +611,7 @@ struct llama_server_context
 
     completion_token_output doCompletion()
     {
-        const completion_token_output token_with_probs = nextToken();
+        auto token_with_probs = nextToken();
 
         const std::string token_text = token_with_probs.tok == -1 ? "" : llama_token_to_piece(ctx, token_with_probs.tok);
         generated_text += token_text;
@@ -1255,7 +1255,7 @@ void beam_search_callback(void * callback_data, llama_beams_state beams_state) {
 struct token_translator {
     llama_context * ctx;
     std::string operator()(llama_token tok) const { return llama_token_to_piece(ctx, tok); }
-    std::string operator()(completion_token_output cto) const { return (*this)(cto.tok); }
+    std::string operator()(const completion_token_output & cto) const { return (*this)(cto.tok); }
 };
 
 void append_to_generated_text_from_generated_token_probs(llama_server_context & llama) {
diff --git a/examples/train-text-from-scratch/train-text-from-scratch.cpp b/examples/train-text-from-scratch/train-text-from-scratch.cpp
index 6fe85d41..947aa7ed 100644
--- a/examples/train-text-from-scratch/train-text-from-scratch.cpp
+++ b/examples/train-text-from-scratch/train-text-from-scratch.cpp
@@ -169,10 +169,6 @@ struct my_llama_hparams {
 
     float rope_freq_base  = 10000.0f;
     float rope_freq_scale = 1.0f;
-
-    bool operator!=(const my_llama_hparams& other) const {
-        return memcmp(this, &other, sizeof(my_llama_hparams));
-    }
 };
 
 struct my_llama_layer {
@@ -929,28 +925,6 @@ void get_example_targets_batch(struct llama_context * lctx, const int * train_sa
     }
 }
 
-
-#ifdef __GNUC__
-#ifdef __MINGW32__
-__attribute__((format(gnu_printf, 1, 2)))
-#else
-__attribute__((format(printf, 1, 2)))
-#endif
-#endif
-static std::string format(const char * fmt, ...) {
-    va_list ap, ap2;
-    va_start(ap, fmt);
-    va_copy(ap2, ap);
-    int size = vsnprintf(NULL, 0, fmt, ap);
-    GGML_ASSERT(size >= 0 && size < INT_MAX);
-    std::vector<char> buf(size + 1);
-    int size2 = vsnprintf(buf.data(), size + 1, fmt, ap2);
-    GGML_ASSERT(size2 == size);
-    va_end(ap2);
-    va_end(ap);
-    return std::string(buf.data(), size);
-}
-
 int tokenize_file(struct llama_context * lctx, const char * filename, std::vector<llama_token>& out) {
     FILE * fp = std::fopen(filename, "rb");
     if (fp == NULL) {
@@ -983,10 +957,10 @@ int tokenize_file(struct llama_context * lctx, const char * filename, std::vecto
     out.resize(size+1);
 
     if (std::fread(buf.data(), size, 1, fp) != 1) {
-        throw std::runtime_error(std::string("unexpectedly reached end of file"));
+        die("unexpectedly reached end of file");
     }
     if (ferror(fp)) {
-        throw std::runtime_error(format("read error: %s", strerror(errno)));
+        die_fmt("fread failed: %s", strerror(errno));
     }
 
     buf[size] = '\0';
@@ -1047,11 +1021,11 @@ void shuffle_ints(int * begin, int * end) {
     if (kid >= 0) { \
         enum gguf_type ktype = gguf_get_kv_type(ctx, kid); \
         if (ktype != (type)) { \
-            throw std::runtime_error(format("key %s has wrong type: %s", skey.c_str(), gguf_type_name(ktype))); \
+            die_fmt("key %s has wrong type: %s", skey.c_str(), gguf_type_name(ktype)); \
         } \
         (dst) = func(ctx, kid); \
     } else if (req) { \
-        throw std::runtime_error(format("key not found in model: %s", skey.c_str())); \
+        die_fmt("key not found in model: %s", skey.c_str()); \
     } \
 }
 
@@ -1136,7 +1110,7 @@ void load_opt_context_gguf(struct gguf_context * fctx, struct ggml_context * f_g
         read_tensor_by_name(opt->lbfgs.lms,  f_ggml_ctx, LLM_TENSOR_OPTIMIZER_LBFGS_MEMORY_S);
         read_tensor_by_name(opt->lbfgs.lmy,  f_ggml_ctx, LLM_TENSOR_OPTIMIZER_LBFGS_MEMORY_Y);
     } else {
-        throw std::runtime_error("unknown optimizer type\n");
+        die("unknown optimizer type");
     }
 }
 
@@ -1315,20 +1289,20 @@ void save_llama_model_gguf(struct gguf_context * fctx, const char * fn_vocab_mod
 
         const int token_idx = gguf_find_key(vctx, kv(LLM_KV_TOKENIZER_LIST));
         if (token_idx == -1) {
-            throw std::runtime_error("cannot find tokenizer vocab in model file\n");
+            die("cannot find tokenizer vocab in model file");
         }
         const uint32_t n_vocab = gguf_get_arr_n(vctx, token_idx);
 
         const int score_idx = gguf_find_key(vctx, kv(LLM_KV_TOKENIZER_SCORES));
         if (score_idx == -1) {
-            throw std::runtime_error("cannot find tokenizer scores in model file\n");
+            die("cannot find tokenizer scores in model file");
         }
 
         const float * scores = (const float * ) gguf_get_arr_data(vctx, score_idx);
 
         const int toktype_idx = gguf_find_key(vctx, kv(LLM_KV_TOKENIZER_TOKEN_TYPE));
         if (toktype_idx == -1) {
-            throw std::runtime_error("cannot find token type list in GGUF file\n");
+            die("cannot find token type list in GGUF file");
         }
 
         const int * toktypes = (const int * ) gguf_get_arr_data(vctx, toktype_idx);
@@ -1356,7 +1330,7 @@ void save_llama_model_gguf(struct gguf_context * fctx, const char * fn_vocab_mod
             // read and copy bpe merges
             const int merges_keyidx = gguf_find_key(vctx, kv(LLM_KV_TOKENIZER_MERGES));
             if (merges_keyidx == -1) {
-                throw std::runtime_error("cannot find tokenizer merges in model file\n");
+                die("cannot find tokenizer merges in model file");
             }
 
             const int n_merges = gguf_get_arr_n(vctx, merges_keyidx);
@@ -1988,7 +1962,7 @@ void opt_callback(void * vdata, float * sched) {
     float min_sched = params->adam_min_alpha / params->adam_alpha;
     *sched = min_sched + *sched * (1.0f - min_sched);
 
-    int impr_plot = std::isnan(opt->loss_after) ? 0 : -(int)(1 + (opt->loss_before - opt->loss_after) * 10.0f + 0.5f);
+    int impr_plot = std::isnan(opt->loss_after) ? 0 : -std::lround(1 + (opt->loss_before - opt->loss_after) * 10.0f);
     printf("%s: iter=%*d, sched=%f loss0=%f loss=%f | improvement: %*d>\n", __func__, 6, opt->iter, *sched, opt->loss_before, opt->loss_after, impr_plot, (int)0);
 
     if (data->shuffle_countdown < n_batch) {